A Method For Mining Key Nodes In Social Networks Based On The Multi-attributes

Along with the vigorous development of online social networks,social networks have been an important platform for people to publish,acquire and discuss up-to-date information.In social networks,influential users have the ability of having an impact on a number of other users with direct or indirect links in a short time,and they also play an important role in the dissemination of information,promotion of new goods and public opinion control.Therefore,the key nodes in the social networks contain great research significance and commercial value.How to judge the actual influence of the nodes in the social network effectively and objectively,and further excavating the key nodes in the network has become a key problem that needs further study.Due to the complexity of social network,some simple mining methods based on neighbors and paths can not be applied to the actual social network.In contrast,the methods based on eigenvector,such as Page Rank?Leader Rank and weighted Leader Rank algorithm,are more meaningful.Thus,this paper analyzes the weighted Leader Rank algorithm in detail,and it is found that while it is applied to the key node mining of social network,the algorithm uses only the topology attribute of the node.As a matter of fact,in addition to the topology attributes,many attributes of the nodes can also reflect the performance of the nodes in social networks,which reflect the recognition level of the other users towards the current user.Consequently,the key nodes mining algorithms should take two aspects into consideration.On the one hand,it is supposed to take the topological properties of nodes into account,on the other hand,it is supposed to combine those properties which reflect the social performance value of the nodes in social networks.In view of the above research,the paper starts with a case study of Zhi Hu network,and analyzes the user's personal attributes and social relations in the network based on the multi-dimensional attributes characteristics of users in social network.Then making using of Zhi Hu network,the paper proposed an improved key node mining algorithm based on weighted Leader Rank.The improved weighted Leader Rank algorithm extends the he comprehensive use of the multi-attribute features of the users,highlighting the average performance of the users in the social network.The paper further extends it into a generic algorithm that makes it suitable for general social networks.In this paper,we first analyze the statistical characteristics of the network,and then compare the node sorting results of the improved algorithm and the original algorithm.Next by referring to the classical literature,the paper designs different experimental verification methods,and analyzes the improved algorithm and the original algorithm respectively from the aspects of validity and robustness.Finally,the improved algorithm is applied to the Github social network,and the two algorithms are experimentally simulated and validated based on the Github social network data.The experimental results show that the improved algorithm is more suitable for mining the key nodes in the social network.The above experimental results show that the improved algorithm can consider both the topological attributes and personal attributes of the nodes in the network,and mine the key nodes that meet the objective needs of social networks.Besides,the improved algorithm compared to the weighted Leader Rank algorithm,in the face of complex network interference edge or interference point shows stronger robustness.